Assessing the Impact of Housing Assistance on Child Health: Does the Local Rental Housing Affordability Context Matter?

Background and Purpose

Federal low-income housing assistance programs represent a potentially powerful lever for driving improvements in child health. However, in the context of a growing nationwide housing affordability crisis, prior research has not fully examined whether and how the impact of these programs varies depending on the local rental housing affordability context. To address this gap, we use linked administrative from the Department of Housing and Urban Development (HUD) and National Health Interview Survey (NHIS) to assess whether the effects of housing assistance on child health differ across communities with different housing affordability contexts. Because housing assistance is likely to affect child health partly via an income effect, we hypothesize that receipt of housing assistance will be more strongly associated with cost-driven health outcomes in areas where renters are more often cost burdened.

Methods

We used the linked HUD-NHIS administrative data for the period from 1999-2018 to identify a cohort of 7,287 children who were residing in HUD-assisted housing at the time of their NHIS interview and a group of 3,262 children who entered HUD housing within two years of their NHIS interview and who thus serve as a pseudo-waitlist control group. We stratify the sample into three groups of children residing in communities with low, moderate or severe rental housing affordability challenges based on the proportion of the renter population in their county of residence that is “cost-burdened” (i.e. paying >30% of their income towards rent). We assess the overall impact of housing assistance on self-rated health, household food insecurity and having any unmet health care need due to cost using regression models that compare children who are current vs. future HUD housing recipients and that adjust for relevant confounders. We estimate a parallel set of regression models among children within each community-level housing affordability strata to assess whether the impact of housing assistance on our outcomes varies depending on the local rental housing affordability context.

Results

We find no overall effect of housing assistance on self-rated health, but find receipt of HUD housing assistance is associated with 28% and 19% reductions in the odds of experiencing food insecurity and having any unmet health care need, respectively. We find no significant association between receipt of HUD housing assistance and any of our outcomes among the groups of children residing in communities with low and moderate rental affordability challenges. By contrast, among children residing in communities with severe rental housing affordability challenges, we find receipt of HUD housing assistance is associated with 33% and 31% reductions in the odds of food insecurity and any unmet health need due to cost, respectively.

Conclusions and Implications

Our results suggest that the positive effects of housing assistance on cost-related health outcomes are most likely to be realized in communities with more serious housing affordability challenges. Amidst growing calls for expansion in the supply housing assistance programs, our findings suggest that geographically targeting such an expansion to communities with high housing costs may yield the biggest overall improvements in child health outcomes.